Don’t Underestimate Cost Modeling
By Justin Fazzari, Ph.D., Senior Broadband Strategist
THERE IS SO MUCH TO DO! Why Go Through the Hassle of Cost Modeling?
- It’s one of the most impactful things you can do to STRETCH your BEAD Allocation
- It’s the most critical part for developing BEAD project areas in the Subgrantee Selection Process.
My house is about 50 years old, similar to most other houses in my neighborhood. One night a couple of months ago my basement began to flood. I called out the plumbing company and they opened the main sewer line between my house and the street and scoped it with a remote camera. They said tree roots had grown into the cracked pipe and they needed to replace the pipe. I quickly went ahead and approved the work, not questioning the price as we had to have it fixed as soon as possible with four people using water in our house.
Once that pipe was replaced, they scoped the second pipe that extends from the edge of the house to the street. It was going to be another several thousand dollars. At that point, the water was draining and so I asked them to stop the work while I got other estimates. I was hoping the other estimates would be less and some were by about three thousand dollars. I then reached out to the first plumbing company and negotiated for a lower price on the second pipe. I ended up saving several thousand dollars by working through these estimates and negotiating with the first plumbing company.
I was talking to my neighbor the other day and he mentioned that he also had to have his sewer line replaced but it backed up when he was out of town. The plumbers did the work inside that was needed, he locked up the house and left for his trip. The same thing happened, once the first pipe was replaced, they found multiple cracks in the second pipe, and it needed to be replaced. Being on a business trip, he went ahead and authorized the additional work right over the phone.
After we compared notes on pricing, I realized that had I not been home and engaged in this process at a very hands-on level, I would have spent about 25% more than I needed to spend to repair my pipes. My neighbor actually ended up spending even more than that.
So, what does this have to do with cost modeling and the subgrantee selection process?
Two Alternatives for Managing the State Broadband Subgrantee Selection Process
In my view, there are two ends of the spectrum for how to conduct the subgrantee selection process:
- Publish a map of your state’s BSL locations and allow ISPs to create their own project areas; or
- Use data analytics to groups BSLs to meet your objectives within your program funding constraints.
Approach #2 attempts to strike a balance between being attractive for ISPs AND including the hardest to reach and most expensive unserved BSLs, thus following the priority requirements of the BEAD program.
On the subgrantee selection side, my neighbor’s approach would be akin to allowing ISPs to create their own project areas with minimal guidelines and parameters for prioritizing unserved locations. ISPs will use their own internal cost model analysis and create project areas with the most profitable outcomes with much less focus on the macro mission of BEAD to provide reliable broadband service at 100/20 to as many unserved and underserved locations as possible. Who could fault ISPs for this approach? It’s all part of their business model. They are accountable to their shareholders to be as profitable as possible. They are doing what good businesspeople are supposed to do.
Like my neighbor signing off on significant construction work without doing market or cost research, provider comparisons and project area negotiations, this problem is magnified if we look at the amount of money states are poised to spend through their BEAD programs.
My approach to the plumbing problem would track similarly to states taking a look at all the factors that go into what makes serving a BSL attractive or not and creating project areas that provide some upside for ISPs but also make sure to fulfill the mission of BEAD.
I basically used a form of cost modeling to save as much money as possible on my sewer pipe while still getting the job completed competently. I asked that the estimate be broken down to show how they arrived at the final cost. These companies provided a document that considered the landscape, distance, general difficulty to install, plus a profit margin built in.
States have similar goals in BEAD to stretch their BEAD allocation as far as possible, to reach as many unserved and underserved locations as possible and if there is money left over, look at the possibilities to fund CAIs and other nondeployment activities.
Factors to Consider for Pre-defining the Project Areas for Subgrantee Awards
If states are engaging in some degree of developing broadband project areas for the subgrantee selection process, they will want to engage in significant cost modeling efforts.
Yes, the CostQuest model data that NTIA has licensed for states to use provides estimates for greenfield prices of providing service to BSLs. However, granular model details are not provided, including key information such as: What are the specific variables? How heavily are they weighed? How well does the model perform? Are confidence levels high or low? Which prices are statistically significant and to what degree? What level of confidence do they have in the model results based on geography such as dense urban areas compared to sparsely populated rural areas?
States can enhance their ability to make their BEAD program a success through data driven cost modeling. This can be done by analyzing additional components to enhance and/or confirm the value of the CostQuest data. Some of these factors include infrastructure distance, passings per mile, outlier analysis, density-based clustering, geographical features, and level of ISP competition.
Through data analytics, confidence matrices can be developed for these influential factors to create BSL profiles. Based on these profiles, BSLs can then be aggregated into BSL project areas that indicate how likely it would be for an ISP to determine this BSL project areas would be favorable investment opportunities to include in their application for BEAD funds.
As another point of reference, States can also take past grant program data such as ARPA and model that through regression analysis, along with other variables to see how the results compare to the CostQuest data.
These approaches (as well as many others that we’ve been exploring) provide a solid foundation for states to begin binning and bundling BSLs into project areas for the subgrantee selection process that makes sense for both ISPs to apply for and to advance the goals of BEAD.
We’ve been using the above strategies to tackle the difficult but important task of figuring out how best to group BSLs so that the states we are working with can get the best “deal” in the subgrantee selection process and STRETCH their BEAD funding as far as possible.
It’s important for states to actively engage in their subgrantee selection process, focusing on effectively managing the development of BEAD project areas and the related broadband infrastructure costs. The approach and methodologies will vary from state to state but the fundamental framework of this data driven approach is the same. It is much like how I obtained additional detailed cost estimates before giving the “go ahead” to my plumbing contractor. Scheduling the time and dedicating effort to do a little information gathering and analysis can improve the outcome significantly.
Join us for our upcoming webinar on Wednesday, November 8, 2023 at 2PM ET: Binning, Bundling and Bargaining: How to Navigate the BEAD Broadband Program Subgrantee Process.